Medicina non fortuna neque personis homines aestimat, uerum aequaliter omnibus implorantibus auxilia sua succursuram se pollicetur.

Medicine does not evaluate persons by their fortune or identity, but promises help equally to all who implore its rescue.

Scribonius Largus, Compositiones, 47 CE1

One of the most traditional difficulties in medicine is to adopt the population and individual perspective simultaneously.2 In proposing a “population view”, that health inequalities are unjust when they result from “an unjust distribution of the socially controllable factors affecting population health and its distribution” (p27),3 Norman Daniels argues for a change in the population end of our thinking about just health. Inasmuch as the socioeconomic determinants of health result in unjust distribution of health, by Daniels’ own definition they affect the distribution of species-typical functioning,4 and thus also negatively impact fair equality of opportunity.5 If we accept this view, changes to public health and public policy are, as he outlines (pp333–51),3 not straightforward. Nevertheless, the way ahead does seem clearer.

What of clinical practice? Restricting the argument to the population perspective may seem to leave this question outside its scope. However, doctors are in fact expected to integrate the population view into their reasoning at the bedside. They do so when they apply population statistics, concern for public health, evidence-based medicine, and their own experience over hundreds, sometimes thousands, of similar cases. They are also simultaneously entrusted with the care of individual patients and with attention to preventive measures which require population-level reasoning. In addition, doctors are faced with the task of setting limits in clinical practice,6 which requires a combination of individual-level considerations and population-level thinking.

This means that reasoning about the socioeconomic determinants of health can, in principle, be applied to clinical care. Hence the following challenge: given what we know about the importance of wealth, education or social status to health, is fairness in medicine served better by remaining neutral to our patients’ social status in setting clinical priorities, or by attempting to equalise unjust health inequalities by giving priority to the socioeconomically disadvantaged at the point of care? Although our initial intuition may be that such a thing would be outrageous, this may not be based on sound justification. The reason may simply be that we have been socialised in the current form of medical fairness, which precludes any priority based on social status. I know that I was. Given the ubiquitous inequalities associated with the socioeconomic determinants of health, however, continuing to attempt neutrality to social status may be intuitive, and nevertheless misguided in the light of recent evidence. In this paper, I will argue that doctors should indeed remain as neutral as possible, and not attempt to give priority to low status. However, recent data on the socioeconomic determinants of health do challenge us to rethink this position, and either to ground it in arguments that take these data into account, or revise it.

“ORDINARY” MEDICAL FAIRNESS

Our starting point is what we could call “ordinary medical fairness”. The current view which can be stated approximately like this: “treat all equally according to their health-related needs”. This can be variously stated as the requirement that “medical care should be distributed on the basis of medical need”,7 or, as stated in the physician’s oath of the World Medical Association: “I will not permit considerations of religion, nationality, race, party politics or social standing to intervene between my duty and my patient”.8 It is quite a high demand. Among other things, it means that doctors are required to set aside considerations of status and “personal value”, even if these happen to be otherwise prevalent in their societies. Though data unsurprisingly suggests that this is incompletely achieved,9 it does seems to be successful at least to some degree.10

The challenge posed to ordinary medical fairness by the socioeconomic determinants of health can be stated as follows. If medicine were deployed on a level playing field as regards disease, then ordinary medical fairness could indeed be our best approach to fostering just health. But bring in data on the socioeconomic determinants of health, and this is no longer the case. We could now reword ordinary medical fairness as follows: “avoid increasing health-related inequalities among those who reach the point of care”. This is much more modest than the previous wording, as it is now explicit that some inequalities will remain despite doctors’ response to medical need. Bring in a concern for the socioeconomic determinants of health, and attempting to reach a fair distribution to health related needs may seem to require much more than that.

FOUR EQUALISATION STRATEGIES

Striving for equality in the delivery of healthcare can take several forms. Starting at the point closest to current practice, and progressing in the order of increased likelihood of controversy, doctors can in theory use at least four strategies in attempting to equalise response to health needs, and ultimately their patients’ health.

Firstly, doctors might deploy efforts to equalise access to healthcare. They could do their best to act against biases built into the healthcare system,11^–14 to fight these biases in their own reasoning,15 16 to the extent that they identify them, and to realise equal treatment for equal need as fully as possible.17 18

Secondly, some interventions, like motivational discussions for smoking cessation, are not equally effective across the socioeconomic gradient.19 To equalise this, doctors could treat patients with similar medical needs differently to reach the same health-related outcome, for example by designing specific strategies for disadvantaged patients. This sort of intervention might also include giving an appointment sooner if we know that this patient had to overcome greater obstacles to arrive at the point of contact with the healthcare system.20

These two strategies afford the same importance to the health-related outcome of interventions for all with similar current need. They present practical difficulties. Implementing them can cost resources (p340)3 and thus give rise to instances of the “best outcomes/fair chances” problem.21 However, they stay within ordinary medical fairness.

Thirdly, however, as some conditions can worsen existing socioeconomic inequalities, treating them can sometimes alleviate these inequalities. This could bolster the indication for the intervention. Treating a child for a hearing impairment will, for example, give her a better chance of academic achievement than she would have had otherwise. Were it not for the possibility of problems in school, an intervention may not be indicated in all such cases. So its indication could increase if the child were otherwise disadvantaged in an area also affected by hearing, such as education, for example if her parents had little education themselves.22 This would be equivalent to focusing on fair equality of opportunity directly, rather secondarily through equal access to healthcare.

Finally, because of what we know about the socioeconomic determinants of health, and inasmuch as we live in a society where they are unfairly distributed, doctors could go one step further and contribute to equalising unjust health inequalities by giving some priority for needed interventions to the socioeconomically worst-off. For example, they could allocate more time in outpatient consultations where trade-offs are often made between different preventive interventions that all take time, or give some priority in waiting lists for elective interventions. A more extreme example might be to use low status as a tie breaker for patients with equal need in situations where for example only one operating room was available.

The latter two strategies give different relative importance to an intervention’s health-related outcome. This importance is subject to the strength of other claims, and these other claims are themselves based on pre-existing inequalities that doctors ought, if we accept these strategies, to decrease. They go beyond duties based on respect for ordinary medical fairness.

Is ordinary medical fairness too modest? In this paper, I focus on the fourth strategy: if medicine is “the ambulance waiting at the bottom of the cliff” (p86),3 does protecting fairness require that doctors give priority to those more likely to fall from the cliff earlier, or harder?

PRACTICAL DIFFICULTIES AND THEIR LIMITS

Although practical difficulties do exist, none of them is insurmountable. Giving priority to low status would require that we be able to identify unjust health inequalities—those stemming from unjust social inequalities. This may be difficult. As we lack broad consensus on whether some types of social inequalities are unjust, it may be restricted to those we can agree on. However, it does seem likely that some would exist. In turn, identifying unjust health inequalities could require identifying causal pathways between unjust social inequalities and effects on health (p102).3 Some are currently insufficiently investigated to be applied to clinical reasoning, but this may be transitory. Nevertheless, as long as some general causal pathways are known, doctors would merely be addressing inequalities incompletely, rather than inaccurately. It would obviously be difficult for doctors admitting a patient with myocardial infarction to their ward to know whether this specific myocardial infarction was caused by socioeconomic disadvantage. This, however, would not be a fatal flaw in the approach as long as it could be identified that some groups started out at a disadvantage, and should be given compensatory priority at the point of care. Finally, as discussed by Norman Daniels,3 giving priority to low status would present clinicians with a version of the unsolved priority problem.21 Equalising outcomes through priority to low status could also sometimes mean that we sacrifice efficacy, making this a form of the “best outcomes/fair chances” problem too. Again, however, this would not resolve whether doctors should give any priority to low status per se. Furthermore, priority to low status may sometimes increase efficacy, as could for example happen if persons of low status were given priority in screening and prevention programmes for pathologies that affect them more frequently. What the compound effect might be is an empirical question.

SHOULD WE GIVE PRIORITY TO LOW STATUS?

Several arguments can be put forward in favour of asking doctors to give priority to low status in the delivery of clinical care. After all, as long as unjust health inequalities exist in our societies, the claim that they require compensation, including through priority in treatment, seems initially at least as convincing as the claim that they require measures aimed at decreasing them at the source. Both strategies share the same outcome as their goal.

Moreover, if the better-off groups were partly responsible for the plight of the less well-off, or had benefited from this plight, then as Daniels states “we should not be complicit in sustaining the advantage they illegitimately acquired by refraining from giving more priority to those they have harmed” (p341).3 As he points out later “Other people also have claims of justice on measures that protect their health” (p343),3 but the trade-off is one that reasonable people will disagree about.

Furthermore, health systems are a space common to all parts of society. One of their roles can be to enhance social cohesion. By giving priority to low status, doctors could compensate the sometimes legitimate impression of being left aside in other social spaces. On the other hand, this could also decrease the role of healthcare systems as spaces where all are treated equally and thus perhaps also their role in enhancing social cohesion.i

In addition, doctors are in a unique position to see unequal treatment,23 and some of the results of socioeconomic inequalities, in their healthcare system. As gatekeepers, they have the means to give priority to some over others.

AGAINST PRIORITY TO LOW STATUS

This last point does lead us to some arguments against priority to low status, only some of which can be answered.

First, this is a proposal for reverse discrimination. Does this mean it is unfair? Not automatically24; and if we accept Daniels’ arguments we have to consider that we do indeed owe everyone an equal chance of being healthy, to a degree that is currently not realised. So such reverse discrimination may be justified. As with all such proposals, it could be difficult to identify who ought to get priority.25 However, studies in the socioeconomic determinants of health have used rather simple variables, so it may not be excessively difficult to determine who should be given priority on the basis of their income quartiles, job and educational level. Another problem is that reverse discrimination can be viewed as retribution, and thus outside the remit of medicine (p342).3 However, even if this argument stands, it will only defeat priority to low status if its intention is indeed retribution, rather than equalisation of health.

Another concern is that ordinary medical fairness fosters a somewhat different value system within medicine as the one outside it. Doctors act as the guardians of an enclosed space where socioeconomic status ought to play no role. There is no inconsistency per se in defending greater egalitarianism within medical care than outside it. It represents recognition of our common biology, and vulnerability to disease. It also recognises that to strive for equality in health is to attempt greater equality of resources, rather than outcomes, as regards our opportunities in life more generally.5 If doctors, however, attempted to compensate for inequalities in the socioeconomic determinants of health, would this not become a form of proselytising? In effect, they would be compensating for social injustices that took place outside the remit of medicine. Inasmuch as they are currently likely to have varying conceptions of what constitute unjust health inequalities, they would also confront the victims of this proselytising with a high risk of arbitrariness in their decisions. It could be objected that this would only be the case were doctors to act on their own. If these were inequities that we as a social group wanted to decrease, however, then we would not only have already adopted a more intelligible view of what these inequities were, but also a more egalitarian value system outside of medicine.

On a more pragmatic note, priority to low status might alienate the powerful. If this happened, the strategy could become counterproductive for the very people it aimed to help. If the typically rather well-off persons in charge of funding decided to give dramatically less to healthcare, the worst off could in the end get less that they currently do even if we gave them priority. This “targeting problem” has been identified in other areas of welfare policy in general, where “optimal policy for the very poor is not necessarily a policy that targets benefits as narrowly as possible once the impact of targeting on political support is taken into account”26 As regards priority to low status in clinical care, there are three possible answers. First, this could be a matter of degree: some priority may be possible without risking a worse outcome for those with low status. Also, it is an empirical question: after all the powerful may still prefer to have a functioning healthcare system even if they get somewhat less priority within it. The degree of priority this would allow may, however, still be insignificant. We should expect the resilience of healthcare systems based on solidarity to grow more fragile as they increasingly become exceptions to broad inequalities accepted in other areas. Indeed, it has been pointed out that healthcare is simultaneously a strongly valued area where we should expect solidarity to be more resilient than in other domains of life, and a “morally more demanding” area where a decline in solidarity would have more impact.27 These, apparently contradictory, characteristics, could suggest that the expectation that medicine will treat all equally—oneself included—bolsters support for collectively funded strictly needs-based healthcare: any weakening of the reciprocity involved could be strongly damaging. Moreover, the impression that those who pay and those who benefit somehow do not belong to the same group would also erode support for healthcare funding. This is indirectly suggested by studies of both political trends28 and individual decisions.29 In such a context, very little may suffice to drive the well-off away, especially at a time of increasingly globalised high-end healthcare offer. The incentive to fund public health systems generously would then decrease among the well-off. Finally, however, this argument rests on the idea that we should do as well as possible by the worst off. Accepting it has other consequences: we ought to channel funding from well-paid and sought-after intervention into coverage for the poor to a greater degree.

A related point is that shaping healthcare delivery in such a counter-cultural way will cost even more resources than the current levelling of social biases. A predictable pitfall would need to be avoided: if we unblind clinical reasoning to social status we may make doctors more likely to follow the biases of their society and give priority to the well-off, rather than the opposite. To the effort required to unblind clinical reasoning would then be added a further effort to help doctors give priority to low, rather than high, status. This makes it more likely that this approach will fail in the balance of the “best outcomes versus fair chances” problem. However, as this problem is still unsolved, the question cannot be closed in this way.

Finally, however, under ordinary medical fairness, we protect two different values by treating all alike: (1) aiming for equal health, and (2) showing equal regard. Norman Daniels speaks of “the neutrality that seems appropriate in medical contexts.” (p334)3 There is value in treating all alike. Data on the socioeconomic determinants of health may shift our view on the first of these goals, but they do not invalidate the second goal. At most, they will force a choice. To justify giving priority to low status, we must give priority to equal health over equal regard in the delivery of care. Deciding which of these values to prioritise would hinge on how important we think each is, but also to what degree it is likely to be affected by giving priority to low status. A further consideration, that to provide equal health is one possible interpretation of what it means to treat people with equal regard, would itself require that the effect be sufficient to aim at equal health credibly. If the resulting compensation was small in terms of health, it might not be worth sacrificing the equal regard shown through neutrality. This would apply if only little priority could be given to low status in practical scenarios, or if priority resulted only in small health benefits for those who received it. This would also apply if the equalising effect was very uncertain. As things stand, we cannot expect the health effects of priority to low status to be large. So I would argue that unless we thought equal regard through neutrality at the point of care to be of very little importance indeed, we should not trade it off by giving priority to low status.

A MORE FUNDAMENTAL QUESTION

Some of these questions rest on a deeper one. Are we aiming for just health, or social justice, or both? If our purpose is more justice in health, and if we consider that social justice is good primarily because it is good for our health, then it may actually not matter so much if we consider that priority to low status is a fair means or not. It would contribute something to a fairer outcome. Equal regard would indeed be of little importance. Doctors, however, would be acting with a rather broader aim than the one they currently endorse. It may be a broader aim than we want to entrust them with, or that they wish to endorse.

If our purpose is social justice through health, however, things may work out differently. The legitimacy of medical intervention would be greater, as would the importance of equal regard. Giving priority to low status may then be far less important than attempting to decrease the negative impacts of current disease on social justice. It could continue to include preventive measures, fair limit-setting and measures to equalise response to similar need, such as designing specific strategies when this is required to reach the same effectiveness, and acting against biases both in doctors’ reasoning and in the healthcare system. Priority to low status, however, could be judged excessive, as maintaining sufficient functioning for a fair range of opportunity may then be considered sufficient without evidence that more health would lead to more fairness for the socially less favoured. It could conceivably still be justifiable in restricted circumstances; perhaps only inasmuch as it could prevent a “double-whammy” effect where disease was expected to deepen an existing social injustice.

We may be aiming at both. Social justice and just health are increasingly shown to be more intricately bound that we ever previously thought. Indeed, if the effects of the socioeconomic determinants of health reflect the way in which we live our very social life and dependence on others,30 then this could even turn out to be a distinction without a difference. Even if this were so, however, it could mean that we cannot lightly make trade-offs for one goal in the name of pursuing the other goal. Giving priority to low status would thus not be justified in this case.

CONCLUSION

Justified claims that we should address the socioeconomic determinants of health could prompt the question of giving priority to patients of low socioeconomic status at the point of care. Although this strategy could go some way towards compensating unjust health inequalities, and the impression of being left aside in other social spaces, in my view even selflessly egalitarian doctors should currently not attempt it. Inasmuch as it may be possible to overcome some of the obstacles presented here, this conclusion is circumstantial. At present, however, such priority would risk being given arbitrarily, could represent a form of medical proselytising, risk leaving the worst-off with less than before, and require teaching doctors to act in strongly counter-cultural ways. Crucially, however, we protect both equal health and equal regard by treating all alike according to need: priority to low status would promote the first somewhat, but at the expense of sacrificing the second. As we currently cannot expect the health effects of priority to low status to be large, this sacrifice seems unwarranted. This does not preclude other measures to equalise response to similar need, such as designing specific strategies when this is required to reach the same effectiveness, and acting against biases both in doctors’ reasoning and in the healthcare system.